AIエージェント
SkillSpector
NVIDIA's security scanner for AI agent skills - check before you install
★ 19.7k スターApache-2.0
ガイドを見る →agent-scanは、snykが公開しているオープンソースのAIエージェントプロジェクトです（GitHub: snyk/agent-scan）。GitHubスターは3,126、フォークは287で、Apache-2.0ライセンスの下で公開されています。GitHubでの最終更新は2026年10月6日です。
Snyk's security scanner for AI agents, MCP servers and agent skills.
Decide the first concrete outcome you want from it, so you can pick the right example, template or mode instead of exploring everything at once.
Clone the repository (or install the published package) and skim its README quick start for the requirements for your platform.
Run the documented install step and set any required configuration — such as provider keys, data sources or local services — before the first run.
Execute the smallest documented example end to end, confirm the output is what you expected, then adapt it to your real data or workflow.
Scan everything, one MCP configuration, one skill, or all Claude skills.
A supported-agents matrix with detection across system, user, project/workspace and extension scopes.
The scanner asks before connecting to MCP servers during analysis.
Findings go beyond inventory, with validation of discovered servers and skills.
A CI mode and inspect-only mode fit pipelines and audits.
Published checksums and a step-by-step verification guide for standalone downloads.
Get the source so you can inspect, configure and run it locally.
git clone https://github.com/snyk/agent-scan.git
cd agent-scanScan your whole machine, or point it at a specific config or skill as the README shows.
uvx snyk-agent-scan@latest # scan the whole machine; see README for targeted scansDownload a release binary, verify its checksum per the guide, and run the same scans.
# download from GitHub Releases, verify checksum, then run agent-scan — see READMEIt is maintained by snyk on GitHub (snyk/agent-scan), where the source, issues and releases live.
Yes — it is free and open source under Apache-2.0, which permits use, modification and commercial redistribution, with licence/notice retention.
No LLM API key is required for scanning; analysis runs against your local configurations as the README describes.
Compared with generic antivirus or config linters, agent-scan's distinguishing focus is: snyk's security scanner for AI agents, MCP servers and agent skills. Pick it when that focus matches your need; the alternative may suit you better if its broader or different scope is the priority.
Clone https://github.com/snyk/agent-scan (or install its published package), follow the README quick start to install and configure it, then run the smallest documented example — the Run section on this page mirrors those steps.
NVIDIA's security scanner for AI agent skills - check before you install
Code, build and evaluate LLM agents with full MCP, Skills, ACP and A2A support.
Simple workflow patterns for building effective agents on MCP.
Open ai agents for mechanical verification for ai coding agents — skills p.
Sync your AI coding setup: skills, agents, rules, MCP, hooks, everywhere.
A Python framework for building MCP servers and agent tools.
LibreAIカタログに掲載されたAIエージェントプロジェクトの中で最もスターの多いもの — agent-scanが合わなかった場合に検討したい、現実的な候補リストです。
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
カタログの中でagent-scanとスター数が近いAIエージェントプロジェクト — 前後の棚もそのまま見て回れます。
OSWorld — benchmark multimodal agents on open-ended tasks in real computer environments.
An auto-company that works 24/7 on your own PC — 14 agent roles, a dashboard, real workflows.
A local MCP server for serious arXiv work — LaTeX section reads, BibTeX and topic watches.
Decentralized P2P mesh for autonomous AI agents — your identity, your devices, no central server
An open protocol for secure AI-driven payments.
Open-source red-teaming framework that stress-tests LLMs and AI agents.